langou

Here are some comments from Sven Hammarling. "Perhaps not surprisingly, I would phrase it the other way round that compiler optimizations should not get in the way of code writers....

> More tests are needed if we want to find those issues with row-major matrices in LAPACKE. Maybe [\LAPACK](https://github.com/tlapack/tlapack) can help testing row-major format for LAPACKE since it is easy...

Thanks @ahnaf-tahmid-chowdhury. This PR is ready for a merge. I will let it sit for a few days. If no one objects by then, I will merge.

Hi Sebastian, I am looking more into this. I only had time for a cursory review at this point. I am reading in the comments of the routine that >...

> "Also, for the DD1 = DD1/GAM**2, instead of dividing by GAMSQ, maybe we should multiply by RGAMSQ." => In the double precision code, RGAMSQ would have to be updated...

I looked into this this morning. Here is a brain dump. I now see that @angsch added a comment. Thanks. I am copy-pasting my brain dump. I'll try to work...

Related. I was reading the code and the WHILE loop looked suspicious. I think you can get an infinite loop if you call the function with DD1 is +Infinity for...

For a run of ssyev or sgeev, what routine do you see? For SYEV, I would expect to see SYMV, SYR2, SYR2K for the symmetric tridiagonal reduction. Then it gets...

For LARFB, you are correct. We could change: "The elements equal to 1 are not stored; the corresponding array elements are modified but restored on exit. The rest of the...

Hi @EduardFedorenkov, > about ORM2R and LARF. Looks like the simple solution is this: > (1) ORM2R: double workspace for ORM2R. One part is for copy of V and another...